Technology Overview
Our solutions support MRCP as an industry standard in speech technologies — making them easy to integrate with most call centers solutions.
Core solutions are based on the latest machine learning algorithms, including DNN (Deep Neural Networks), RNN (Recurrent Neural Networks) and Support Vector Machines.
Our system works with extra-large vocabularies in real time. Vocabulary, acoustic and language models are pre-compiled into one static search space.
Languages
We now support English (US & UK), German, Swiss-German, French, Italian and Russian.
New languages are being added in line with market trends.
Technologies
Spitch SignyFi
The Spitch semantic tagging and inference solution offers a high fidelity answer to all of your company's information needs.
Spitch VeryFi
Spitch provides a highly competitive, customizable Speaker Verification suite that can improve the security of your business with a minimum interaction from the user side.
Spitch CodyFi
Spitch ASR offers solutions supporting customizable expert grammars, dialplan design and implementation, domain adaptation and general speech-to-text conversion.
Hardware
X86 architecture.
Currently supported OS — Red Hat Enterprise Linux 64-bit and Ubuntu 64-bit.
We can support any other Unix-based OS if required.
Back-end
We provide the back-end technologies that are responsible for the critical speech-to-text, semantic interpretation, and speaker verification processes.
User access is through a customized frontend of the existing Call Operator Solution (e.g. Call Management System, Call Recorder System, Quality Management System, Customer Relationship Management) or via a dedicated user interface which can be developed based on the customer requirements.
Our back-end solutions include all components needed for deployment in an HA architecture. The particular implementation scheme used depends on the customer HA architecture and can be designed at the development stage of a project.
Call Us
